[bodini]

From the G320 manual:

"If the lead length is more than 18” use a 1000 uF capacitor across the G320 power supply terminals."

Sorry, I am am uneducated in this electronics stuff. My leads are about 15' (feet) and the power supply is 90v.

So that requires a capacitor whos nominal voltage is higher than the power supply? (1000 uF - 100v in my case)

And said capacitor is to be mounted in the Gecko across pins 1 & 2, paying attention to polarity of the capacitor?

Thanks.

[Mariss Freimanis]

The maximum voltage for the G320 is 80VDC. It will work at 90VDC but you have very little safety margin left when you use 90VDC.

Mariss

[bodini]

Ok, so I went a had a better look at the power supply and it is actually 70V. The label was partially obscured when I first posted, so I had a real good look after the comment by Mariss and sure enough it was 70v. This machine is a retrofit and roughed out by a very experienced CNC guy, so I figured I must be in the wrong, which I was.

But anyway, back to the initial question. Due to no objections voiced as of yet, I assume a 100v 1000uf capacitor is the right way to go.

[morestuff]

Yes, a 100 volt, 1000 uF capacitor connected to the power
terminals of the Gecko drive with short wires is a good way
to go.

Regards,
Steve Stallings
www.PMDX.com

[ptwysocki]

Steve,
Would you happen to know what advantage does the capacitor across the power leads do, particularly in this situation, (70vdc to gecko CNC sys)? Is it for power swings, or noise, etc. What problem would you expect to see or have without one installed?

Just wondering, Thanks
Paul

[Richards]

A DC power supply usually consists of three pieces, minimum:

1. Transformer

2. Bridge rectifier

3. Capacitor.

The transformer changes the amplitude of the voltage from the "line voltage" available to the desired output voltage.

The Bridge rectifier gives full-wave rectification, meaning, for all practical purposes that the A.C. sine wave is changed to a positive voltage only sine wave with 2X the frequency and 1/2 the total amplitude of the A.C. sine wave. In other words, A.C. line voltage in the U.S. is really about +170V to -170V when measured with an oscilliscope with a period of 1/60th second per sine wave. After rectification, that would look like a +170V at 1/120th second per "wave".

After rectification, the "wave" is not smooth DC voltage, but a very choppy half-wave that ranges from near zero-volts to the maximum voltage output by the transformer.

A capacitor "filters" those waves, i.e. stores power between cycles so that the power looks and acts like D.C. with very little ripple. If the capacitor is too small, you will have excessive "ripple". Unless the capacitor is extremely oversized, no harm will be done. An excessively large capacity capacitor can look like a momentary short when the power supply is first turned on, so I like to use a capacitor that is at least 80% of the desired value, but not more than 150% of that value.

Putting a capacitor on the motor leads helps filter out ripple caused by the voltage "used up" by the wire's resistance. Feel the temperature of the wires. If they are warm, they have enough resistance to cause a voltage drop. The amount of that drop multiplied by the current being drawn is the Watt level. Watts = heat. Use larger gauge wire to decrease the resistance and the heat. Putting a capacitor at the motor smooths out the ripple that results from the additional current "used up" or wasted by the long leads.

[morestuff]

The G320, and also the stepper drivers for that matter, regulate
the current in the motor by rapidly switching the voltage applied
to the motor. This power is pulled from the power input in current
pulses, and also sometimes returned to the power supply in pulses.

These amplitude of the current pulses can be significantly higher
than the average current being drawn. For this reason it is important
to provide a low impedance source for the current supply. The driver
itself has a filter capacitor inside to help with smoothing out these
current pulses, but it also relies on the capacitors in the power supply
to a large degree.

If the wires between the driver and the power supply are long
enough to have added impedance, then the capacitors in
the power supply will be less able to assist in the task. The
result is additional stress on the capacitor inside the driver
and more voltage ripple in the supply seen by the driver. By
adding another capacitor near the driver you bypass the extra
impedance of the wiring and once again provide a way to share
the burden of filtering the current pulses required by the driver.

Regards,
Steve Stallings
www.PMDX.com

[Al_The_Man]

When adding Extra capacitance, it is advisable to keep the value as low as possible, especially if the P.S. is sized close to the required current demand.
A little known fact is that any increase in capacitance will affect the VA requirements of the P.S.
Al.

[ptwysocki]

Thanks for all the information,,, now I think I have a problem! ha

Im a novice in electronics, but can follow instructions. As with many on CNC Zone, we appreciate all the information and education you all provide.

When adding capacitors to the Gecko 320x, Gecko recommends 1000u Caps installed across power supply when over 18" long. What about adding additional Caps to the "long wires" to the Servo motor (armature+-)? Is this redundant, helpful or harmful? If helpful, do you install the Caps closer to driver or motor and is the Cap size the same?

Again thanks.
Paul

[morestuff]

Do NOT try to put a large capacitor across the armature connection
either at the drive or at the motor. This is because the signal being
generated by the G320X is a PWM signal that rapidly switches the
full power supply voltages and a large capacitor will draw a huge
amount of current trying to smooth it out. The G320X is designed
to drive an inductive load such as motor windings.

Sometimes very small, suitably rated high frequency capacitors
are put across the armature terminals at the motor to filter out
interference from brush noise, but I do not recommend it for
servo motors.

Regards,
Steve Stallings
www.PMDX.com

[Mariss Freimanis]

I'll reinforce what Steve just said. Never put any components on the armature leads no matter how long they are. No resistors, no capacitors, no MOVs, no fuses, no nothing. Nothing but wire from the drive all the way to the motor.

Let me introduce myself just so you know who I am. I design all the drives at Geckodrive so you can know the advice I give is solid.

The purpose of the "remote" capacitor at the drive is to restore the power supply input to its full vigor. It's not quite what it was after it has traveled a few feet of wire from power supply to the drive. The wire resistance voltage drop has taken its toll on how much peak current it can deliver during acceleration or deceleration. The capacitor restores the power supply because it acts as a flywheel. It stores up energy that can quickly delivered to the drive right then and there when it needs it.

Think of it as a vitamin supplement.

Mariss

[ptwysocki]

Mariss, Steve,

Thanks for the advise and the clear understanding why. Im trying to track down what I think is noise problems on my system, thats why I was intrested in capacitor info. I think its the 70VDC power supply.

One quick question for the Gecko experts, is it common to have the 320x PID tuned up "only" an 1/8 of a turn (maybe less) off full CCW (~8 oclock). any further than an 1/8 turn on P I or D, and I get lots of servo grumbling. I tune by ear, and servos run fine, but it contradicts all other instructions ive read (as in, usually start at 11oclock and go CW). And its a very small window to get the motors quiet and smooth. I plan on adding 1000u 100v caps as recommended by Gecko, will this help? The power leads were never more than 24" long.

The power supply IS next to the C11 BOB / 320x "bay" separated by aluminum partition, maybe 8" away, I was trying to keep my 70volt supply close to 18" leads to the 320x's. Maybe too close? I plan to relocate the PS temporarily far away on the floor as a test.

COMPONENTS:
USA 220vac - 70vdc PS
320X
1100oz Servos 3 axis plus slave (4)

Any ideas why the PIDs only have an 1/8 turn before noise or fault?
Is the PS noise a possibility?
Thanks Paul

[keith20mm]

The maximum voltage for the G320 is 80VDC. It will work at 90VDC but you have very little safety margin left when you use 90VDC.

Mariss

Mariss, my power supply is measuring 82VDC. Is it possible to use the three G320 servo drivers or will I have to reduce voltage somehow?

I ask because you indicate above that "you have very little safety margin left when you use 90VDC"..

Is there a limit pot I can increase?